Johnson Controls International plc (NYSE: JCI) — A Genuine Turnaround, Priced as Though It Is Already Finished

Johnson Controls International plc — NYSE: JCI — CIK 0000833444 — GICS Industrials / Building Products Price reference: $144.96 (close 2026-06-12) · Market cap ~$88.6B · Enterprise value ~$98B · Shares ~611M Fiscal year: ends September 30 · Domicile: Cork, Ireland (plc; US SEC filer) · Employees: ~94,000 Date of memo: 2026-06-13 · Fresh coverage

⚡ Claude’s Take

This block is the author’s own independent, subjective opinion and general information only — it is not investment advice and not a recommendation to buy or sell any security. The analysis that follows takes no position, carries no recommendation, and contains no price target except where this block explicitly states one.

Verdict: HOLD / “right turnaround, full price.” Accumulate only on a meaningful de-rating — a buyer’s zone opens roughly in the $110–$125 area (~23–26x forward adjusted EPS). At ~$145 (~30x forward / ~22x EV/EBITDA / a ~2.4% free-cash-flow yield, and the 98th percentile of its own history on price-to-sales) the stock already pays for the convergence to Trane-like economics before that convergence shows up in the returns. Not a short — the operational momentum is real and the order book is at a record.

Johnson Controls is a genuinely better company than the sprawling, under-managed conglomerate it was three years ago. It has exited residential/light-commercial HVAC (sold to Bosch in 2025), collapsed four segments into three, installed a Danaher-bred CEO (Joakim Weidemanis) running an explicit Danaher-style “business system,” and is converting a data-center cooling boom into 30–40% order growth and a record backlog. The numbers are inflecting the right way: adjusted EBIT margin reached 15.5% in the March-2026 quarter (+310bps YoY), adjusted EPS grew 45%, and management raised FY2026 adjusted EPS guidance to ~$4.85 (~+30%). This is a real self-help story with a long runway — JCI’s margins and ~10% ROIC sit roughly half-way between where they were and where best-in-class peer Trane already operates (~20% applied-segment margins, ~28% ROIC). The gap is the opportunity.

My problem is entirely the price. The stock has already tripled off its 2023 lows (~$50 → ~$145), and at ~30x forward earnings it trades within a whisker of Trane’s multiple while earning roughly half of Trane’s returns and carrying a heavier legacy tail (asbestos, environmental, and open-ended AFFF/PFAS litigation from the Tyco lineage, plus an Irish-inversion tax structure). The market is capitalizing the turnaround as if it were substantially complete. That is the same trap the peer set keeps setting: pay a quality multiple for quality-in-progress and you inherit all the de-rating risk if execution stutters, with little upside left if it doesn’t. The framing is operational-turnaround-priced-for-completion — admire the trajectory, demand a better entry. Conviction: medium. The single piece of evidence that would flip me bullish: two-to-three more quarters of segment-margin and ROIC expansion toward high-teens/low-20s with the data-center backlog converting on schedule — proof the Danaher system is structural, not cyclical — ideally bought after a multiple reset. The single piece that would flip me bearish: the data-center order surge (orders +30–40%) normalizing into a 2027 air-pocket while margins plateau near current levels, against a top-of-group multiple. Tag: “The Danaher playbook, bought at the Danaher price.”

1. Executive Summary

Johnson Controls International is, after a multi-year portfolio teardown, a pure-play commercial-buildings company: HVAC (chillers, applied/airside equipment, heat pumps), fire detection & suppression, security, and building controls/automation (Metasys, OpenBlue), sold and serviced across ~150 countries. Following the July-2025 divestiture of its residential & light-commercial HVAC business to Bosch (~$5.6B net cash; ~$8.1B headline enterprise value) and a string of smaller exits (Air Distribution Technologies in 2024, ADT Mexico residential security in 2025), continuing operations generated FY2025 revenue of $23.6B (+3% reported, mid-single-digit organic) across three regional segments — Americas (~67% of sales), EMEA (~21%), and Asia Pacific (~12%). Roughly 68% of revenue is products & systems (install) and 32% is services; the US is ~56% of sales.

The investment story is an operational turnaround colliding with a full valuation. In March 2025, JCI installed Joakim Weidemanis, a long-time Danaher executive, as CEO, and he is deploying a textbook Danaher-style operating system (“simplify / accelerate / amplify,” 80/20, kaizen, daily management, lighthouse sites). The early financial signature is encouraging: gross margin has climbed from 33.2% (FY2020) to 36.4% (FY2025); GAAP operating margin from 7.8% to 12.0%; FY2025 incremental operating margin was 64%; and the most recent quarter (Q2 FY2026, reported May 2026) printed a 15.5% adjusted EBIT margin (+310bps), 45% adjusted-EPS growth, orders +30% (on top of ~+40% the prior quarter), and a record backlog management pegs at ~$20B (+26%). Management raised FY2026 adjusted-EPS guidance to ~$4.85 (~+30% YoY) on ~6% organic growth and ~50% operating leverage. The growth engine is unmistakably data-center thermal management — York chillers, the Silent-Aire air-handling franchise, and a nascent coolant-distribution-unit (CDU) business — supplemented by the recent Alloy Enterprises acquisition for advanced heat-transfer technology.

The franchise quality is real but second-tier within its own peer group. JCI sits at the bottom of the margin and return hierarchy among the scaled commercial/applied-HVAC majors: a ~10% ROIC versus Trane’s ~28% and Lennox’s ~33%, and segment margins (17.0% blended EBITA) below Trane’s ~20% Americas operating margin. Its moat — an installed base plus controls plus a multi-year service book on mission-critical building systems — is genuine in concept but less monetized than Trane’s (JCI’s services are 32% of revenue and the company does not disclose service gross margin or attach economics with Trane’s clarity).

Valuation is the crux. At ~$145, JCI trades at ~30x forward adjusted EPS, ~22x forward EV/EBITDA, a ~2.4% free-cash-flow yield, and — most tellingly — the 98th percentile of its own ten-year history on price-to-sales and price-to-book (the P/E percentile is artificially low because trailing GAAP EPS is inflated by a ~$1.5B one-time Bosch divestiture gain). The market is paying a near-Trane multiple for a still-converging business. The embedded expectation is that the Danaher transformation will substantially close the margin/ROIC gap to Trane and that the data-center cooling cycle is durable rather than a pull-forward. Both may prove true — but at this price there is little margin of safety if either disappoints. No price target and no recommendation appear below this section; the body discusses valuation only as embedded expectations and scenarios.

2. Business Overview

Johnson Controls is one of the oldest names in building technology — founded in 1885 by Warren Johnson, inventor of the electric room thermostat. Its modern shape is the product of two transformational events: the 2016 merger with Tyco International (which moved the legal domicile to Ireland, added fire & security, and created the legacy-liability tail discussed in §9), and the 2016 spin-off of its automotive seating/interiors business as Adient. Over 2021–2025 the company completed its metamorphosis from an industrial conglomerate (automotive batteries — Power Solutions, sold to Brookfield in 2019; residential HVAC) into a focused commercial-buildings platform.

What it sells. JCI designs, manufactures, installs, and services the systems that heat, cool, ventilate, protect, and run commercial and institutional buildings:

• HVAC & controls (the core): applied chillers (the York brand), air-handling units (including the Silent-Aire data-center franchise), commercial heat pumps, building-management/automation systems (Metasys), and the OpenBlue digital/AI software layer. This is the largest and fastest-growing piece, anchored by the data-center, healthcare, pharma/biologics, semiconductor, and advanced-manufacturing verticals.
• Fire detection & suppression: alarm panels, detectors, sprinkler and suppression systems (the Tyco/Simplex/Autocall heritage).
• Security: access control, video, and integrated electronic security for commercial customers.

How it makes money. Revenue splits into two economic engines:

1. Products & systems / install (~68% of revenue): equipment sales and engineered installation projects — longer-cycle, backlog-driven, project-margin businesses. This is where the data-center order surge lands.
2. Services (~32% of revenue): inspection, maintenance, monitoring, repair, retrofit, and replacement under multi-year contracts on an enormous installed base. This is the recurring, higher-quality annuity — the “razor-blade” attached to decades of installed equipment, and the strategic priority of the prior (Oliver) and current (Weidemanis) management teams alike.

Segments (post-April-2025 reorganization, continuing operations, FY2025):

The Americas is the profit core (NA commercial/applied HVAC + service attach is the best pool in the industry); EMEA is the structural laggard (~300–400bps below regional peers, a fixable self-help target); APAC carries China weakness offset by Southeast-Asia and data-center demand.

Verdict: A focused, scaled, mission-critical buildings franchise with a genuine recurring-service layer — a fundamentally better and simpler business than the conglomerate of three years ago. The model is sound; the question the rest of this memo presses is quality and price, not viability.

3. Industry Dynamics

Structure. Commercial and applied HVAC is, at the high end, a consolidated oligopoly — Trane, Carrier, Johnson Controls, Daikin, Mitsubishi Electric, and Bosch — sitting atop a fragmented, distributor-and-contractor-mediated low end. The top players collectively hold only ~40% of commercial-applied revenue, but the engineered-systems segment (large chillers, central plants, building-controls integration) is structurally concentrated and high-barrier. Fire & security is similarly a scale game layered onto code-driven, recurring inspection revenue. This is, by the standards of industrials, a structurally attractive industry: high barriers to entry (engineering depth, brand, code/efficiency regulation, dealer/contractor lock-in, installed-base service density), rational competition among the scaled majors, and a long secular demand backdrop.

The profit pool sits in North American commercial/applied HVAC plus attached services — precisely JCI’s Americas segment. This is where incumbents earn ~18–21% segment margins and high-teens-to-mid-20s ROIC. The weaker pools (residential, which JCI has now largely exited; lower-end security service) are commoditized and cyclical.

The data-center demand thesis is the dominant near-term driver and the spine of the bull case across the whole peer group. The numbers are large: third-party sizing puts data-center cooling TAM at ~$26B (2025) growing to ~$128B by 2033 (~22% CAGR), with the liquid-cooling subset compounding even faster as next-generation GPU racks (~250 kW and rising) force the transition from air to liquid cooling. Hyperscaler capex — the top four spent ~$413B in 2025, with more committed for 2026 — flows disproportionately to power and thermal management. JCI is a credible front-line beneficiary: its York chillers and Silent-Aire air-handling units are designed into “AI factory” reference architectures, and management reports the air-handling fear (that liquid cooling would cannibalize AHU content) has proved overdone — non-chip heat sources keep AHU content growing, while the CDU business is just beginning to ramp (~$100M expected in FY2026 against a “hundreds of millions” pipeline).

The capital-cycle caveat (Marathon lens) is the single most important industry risk. Data-center cooling is the one pocket where very high incumbent returns are actively pulling in capital: Vertiv, Schneider, Carrier, JCI, AAON, and a host of liquid-cooling specialists are all racing in. High returns that attract a flood of capacity tend to mean-revert. The right framing is that data-center cooling is high-growth optionality, not a durable moat — its scarcity margins are more likely to compress over 3–5 years than to hold. A booking surge with book-to-bill far above 100% is a demand cluster, not a new permanent run-rate, and carries 2027 normalization risk.

Secular drivers, pressure-tested. Electrification/decarbonization of heat (heat pumps), tightening energy codes (which raise unit content and entrench engineered-systems incumbents), refrigerant transitions (the AIM Act phase-down), indoor-air-quality awareness, and the replacement of an aging installed base are all genuine multi-year tailwinds. But they are slower and lumpier than the narrative — US a/c and heat-pump shipments actually fell ~20% in 2025 on destocking and pre-buy payback, and the residential energy-credit repeal removed a subsidy crutch. For JCI specifically, the commercial/institutional and data-center exposure insulates it from the residential air-pocket better than most.

Verdict: Structurally good industry — concentrated at the profitable high end, high barriers, real secular tailwinds. But the specific growth leg the market is capitalizing (data-center cooling) is exactly where the capital cycle is turning hostile, which argues for treating today’s hyper-growth orders as cyclical optionality rather than a permanent moat.

4. Competitive Position

The moat — real, but second-tier. JCI’s competitive advantage is a demand-side customer-captivity / switching-cost moat, reinforced by scale and brand. The mechanism: once a York chiller plant, a Metasys building-management system, or a Simplex fire network is installed in a mission-critical building (a hospital, a data center, a semiconductor fab), the cost, risk, and disruption of switching vendors for service, parts, retrofits, and controls integration is high. That captivity is monetized through a multi-year service annuity on a vast installed base, and is deepened by the OpenBlue digital layer (which makes the building’s data and optimization JCI-native). On the product side, management makes a credible vertical-integration argument in chillers: JCI designs and manufactures its own compressors, variable-speed drives, magnetic (oil-free) bearings, heat exchangers, and embedded controls — over 1,000 patents across the five subsystems — where much of the industry relies on third-party compressor platforms. That ownership enables application-specific performance and a structural cost advantage in the high-capacity, high-precision applications (data centers, biologics) that are growing fastest.

Where it shows up — and where it doesn’t. A moat must surface in financial outcomes. JCI’s do, partially: ROIC has roughly doubled from 5.7% (FY2020) to ~10% (FY2025), gross margin has expanded ~320bps, and the service book grows mid-single-digit-plus. But by the test that matters — does the advantage produce best-in-class economics? — JCI lags the peer set it most resembles. The evidence is consistent: JCI’s margins run meaningfully lower than Carrier, Trane, and Lennox. On the head-to-head metrics:

The structural read: JCI owns Trane-adjacent assets (the same Americas commercial/applied profit pool, a comparable chiller franchise, the same data-center adjacency, a service book of comparable relative size at 32% of revenue) but earns Carrier-or-worse current economics (~10% ROIC, lower margins). It is ahead of Carrier in service-attach orientation and is a cleaner pure-play than Carrier (which diluted itself with sub-9%-margin European residential heat pumps via Viessmann and carries goodwill-impairment risk). It is well behind Trane on the two metrics that define the moat — margin and return on capital.

The bull’s anchor is precisely this gap: JCI has Trane’s raw material at sub-Trane economics, and a CEO hired to run the Danaher playbook that closes such gaps. If even half the distance to Trane’s ~20% margins / ~28% ROIC is traveled, the earnings base expands materially. The bear’s anchor is that the gap is not yet closed — the superior economics are a hypothesis (a turnaround thesis), not yet evidence, while the stock is priced almost as if they were.

Verdict: A durable but second-tier moat — genuine customer captivity and a real (if under-monetized) service annuity, with credible chiller vertical-integration, but returns that sit at roughly the cost of capital plus a modest spread, not the wide spread Trane earns. The competitive position is improving, not dominant.

5. Growth History and Forward Opportunities

History. On a continuing-operations basis, revenue has been choppy as the portfolio was rebuilt: $22.3B (FY2020, restated) → $23.7B (FY2021) → $20.6B (FY2022) → $22.3B (FY2023) → $23.0B (FY2024) → $23.6B (FY2025). The headline volatility reflects divestitures and resegmentation; underlying organic growth ran mid-single-digit-to-high-single-digit through the post-COVID commercial-construction and service recovery (FY2023 organic +8%, service +10%), interrupted by a September-2023 cybersecurity incident and 2024 China/residential weakness. The more durable story is margin-led earnings growth: adjusted EPS compounding in the mid-to-high teens as productivity, price/cost, and mix improved.

The current inflection is order- and backlog-driven. The data-center wave has produced an unusually strong demand signal:

• Orders +30% in Q2 FY2026 (Americas +40%, led by ~60% systems growth on large-scale data-center projects; EMEA +11%; APAC +4%), on top of ~+40% the prior quarter.
• Record backlog of ~$20B (+26% YoY), of which management expects ~70% to convert to revenue within 12 months (the balance gated by customers’ power/electrical-infrastructure timing). (Note: the FY2025 10-K disclosed a “Building Solutions” backlog of $14.9B at 9/30/25 and remaining performance obligations of $22.7B; the ~$20B figure is management’s current backlog metric as of Q2 FY2026 — a definitional step-up worth tracking, but directionally a genuine record.)
• Revenue growth ~6% organic in the most recent quarters — notably lower than order growth, which is the crux: orders and backlog are running well ahead of deliverable revenue, creating both forward visibility and 2027 normalization risk if the order surge does not sustain.

Forward opportunities:

1. Data-center thermal management — chillers (air- and water-cooled), Silent-Aire AHUs, and the ramping CDU/liquid-cooling business (Alloy Enterprises adds proprietary heat-transfer/material-science capability). The richest growth vector, but also the one most exposed to the capital cycle (§3).
2. Service-attach expansion — converting the large install base (and the surge of new chiller installs) into multi-year service contracts. Management cites a West-Florida pilot that tripled service agreements following new-chiller commissioning by redesigning the sales process — exactly the kind of process-driven, replicable improvement the business system is meant to scale. This is the highest-quality growth lever because it compounds recurring, higher-margin revenue.
3. EMEA and fire/controls margin catch-up — applying the operating system to under-managed regions and the less-differentiated (security, fire) franchises, where management explicitly sees “as big or bigger” opportunity than HVAC.
4. Adjacent verticals — biologics, semiconductors, advanced/battery manufacturing, healthcare: energy-intense, precision-cooling end-markets where JCI’s application-specific systems command premium economics.

Verdict: High-quality growth in mix, with a cyclical asterisk. The service-attach and margin-catch-up levers are durable and self-help-driven; the data-center order surge is genuine but partly cyclical and partly a pull-forward. The growth is real — the question is how much of it is permanent versus borrowed from 2027.

6. Financial Quality

Revenue & margins. FY2025 continuing revenue $23.6B with a clearly improving margin structure: gross margin 36.4% (vs 33.2% in FY2020), GAAP operating margin 12.0% (vs 7.8%), and a 64% incremental operating margin — the signature of operating leverage finally being captured. Segment EBITA margin was 17.0%, and the Q2-FY2026 adjusted EBIT margin of 15.5% (+310bps YoY) shows the trajectory continuing into the current year. This is the single most important fact in the financials: the turnaround is visible in the margin line, not just the narrative.

Earnings quality — read continuing, adjusted, not GAAP-total. FY2025 reported diluted EPS of $5.03 is materially flattered by a one-time item: the Bosch divestiture produced a ~$1.5B after-tax gain recorded in discontinued operations ($2.40 of the $5.03). The relevant run-rate figures are:

• GAAP continuing EPS: $2.63 (depressed by ~$546M restructuring/impairment and elevated transformation/corporate costs).
• Adjusted continuing EPS: $3.76 (+17% YoY) — the right base.
• FY2026 guide: ~$4.85 adjusted (~+30%), aided by both margin expansion and the share-count reduction from the $5B ASR.

This is a quality-of-earnings flag the market must navigate: trailing GAAP EPS looks high (inflating the apparent cheapness of the P/E and depressing the AZI P/E percentile to the 36th), while the honest, continuing-operations earnings power is the ~$3.76 → ~$4.85 path.

Cash flow. Continuing-operations free cash flow was ~$2.1B in FY2025 (operating cash flow ~$2.55B less ~$434M capex). Capex is light at ~1.8% of sales — an asset-light field-and-service model. Management guides to ~100% adjusted FCF conversion for FY2026. One caveat: FY2025 reported operating cash flow was muddied by ~$1.16B of discontinued-operations outflows; the underlying continuing-ops conversion is solid but the multi-year cash-flow series requires care to normalize.

Returns on capital. ROIC ~10% (FY2025), up from 5.7% (FY2020) — improving but unremarkable, and roughly half of Trane’s. ROE is not a usable metric here: the reported FY2025 ROE of ~764% is an artifact of the ASR shrinking book equity to $12.9B (from $16.1B), not a sign of profitability. Goodwill ($16.6B) exceeds total equity, so reported ROE and book-value ratios are distorted; ROIC and EV-based multiples are the honest lenses.

Balance sheet. Total debt $9.9B, cash $0.4B, net debt $9.5B (~2.0x EBITDA, within the company’s target range). The ASR mechanically lifted net-debt/net-cap to 42.4% (from 35.6%) by shrinking equity, but leverage is moderate and the maturity ladder is termed-out at mostly low fixed coupons. Pension is essentially fully funded (net obligation ~$211M — immaterial). The real balance-sheet tails are off the operating ledger: legacy asbestos (net ~$48M), environmental reserves ($187M), AFFF/PFAS litigation (open-ended; see §9), and $1.9B of unrecognized tax benefits tied to the Tyco/Adient separations and Irish domicile.

Dilution / SBC. Stock-based comp is modest (~$140M, <1% of revenue) — far lower intensity than software/medtech names — and the share count is falling (693M → 612M over FY2024→FY2025 via the ASR), so dilution is not a concern.

Verdict: Economics are improving with scale — the margin and incremental-margin trajectory is the strongest evidence for the bull case, and the cash model is clean and asset-light. But returns are still only modestly above the cost of capital, GAAP-total earnings overstate run-rate power, and the balance sheet carries legacy tails that the operating metrics don’t show. Quality is rising from a mediocre base, not yet high.

7. Capital Allocation

The framework. Under the prior CEO, JCI’s stated policy was to return ~100% of free cash flow to shareholders via dividends and buybacks. The Weidemanis era inherits a more flexible posture: invest in the business (capacity, the operating system, bolt-on technology like Alloy), maintain the balance sheet (~2x net leverage), and return the balance.

Track record (FY2021–FY2025):

• The $5.0B accelerated share repurchase (ASR) funded by Bosch proceeds is the headline FY2025 capital action: launched 8/7/2025, fully paid 8/11/2025, with an initial delivery of 43.1M shares and final settlement in Q2 FY2026. It reduced the share count ~7% and is the single biggest driver of the FY2026 EPS-growth optics. ~$4.8B remains on the authorization. Assessment: returning divestiture proceeds via buyback is defensible, but the timing — buying back ~$6B of stock with the shares near all-time highs (and a 98th-percentile price-to-sales) — is not obviously value-accretive. It flatters per-share metrics while paying a rich price for the company’s own equity.
• Dividend grows slowly (~2%/yr; ~30–40% payout on adjusted continuing EPS; ~1.0% yield). A low-yield, modest-growth policy — secondary to buybacks.
• M&A has been net divestiture (Bosch, ADTi, ADT Mexico) — a deliberate, value-rational simplification — supplemented by small, capability-focused tuck-ins (Alloy Enterprises). This is the right shape: shed the commoditized/cyclical, buy targeted technology.

The incentive red flag. The executive compensation plan rewards growth and cash conversion but not capital efficiency. The annual bonus (AIPP) keys on three equally-weighted metrics — EBIT growth, revenue growth, and FCF conversion (FY2025 paid out at 176.6% of target); the long-term PSUs vest on cumulative pre-tax earnings, recurring revenue, and relative TSR vs the S&P 500 Industrials. There is no ROIC or return-on-capital metric anywhere in the plan. For a company in the middle of heavy M&A/divestiture churn, a Danaher-style “operating system” pitch (Danaher itself is famously ROIC-obsessed), and a sub-peer ~10% ROIC, the absence of a capital-efficiency gate is a genuine governance weakness — it incentivizes growth and EPS (achievable via buybacks and acquisitions) over the return on capital that would actually validate the turnaround. The new CEO’s $5M inducement award is equity (not a cash make-whole), which is alignment-positive; but insiders as a group own <1% of shares, so alignment runs through unvested grants, not skin in the game.

Verdict: Mixed-to-improving. The portfolio simplification is intelligent and value-rational; the balance sheet is prudently managed; SBC is low and the count is falling. But the $5B buyback at a record multiple is questionable value, the dividend is an afterthought, and the incentive structure conspicuously omits return on capital — the one metric a credible Danaher-style transformation should be judged on.

8. Changes and Headwinds — Last Two Years

Strategic / portfolio (transformational):

• CEO transition (March 2025): George Oliver → Joakim Weidemanis (ex-Danaher EVP). The most important change — it reframes JCI as an operating-system turnaround. CFO Marc Vandiepenbeeck (appointed 2024) remains.
• Bosch divestiture (closed July 2025): exit of residential & light-commercial HVAC (including the 60%-owned Hitachi JV) for ~$5.6B net; the act that completed the pure-play-commercial pivot.
• Resegmentation (April 2025): four segments (Global Products, BSNA, BS EMEA/LA, BS APAC) collapsed into three regional segments (Americas, EMEA, APAC) — simpler, regionally accountable, and aligned to the new operating model.
• Smaller exits: Air Distribution Technologies (2024); ADT Mexico residential security (2025, held-for-sale).
• Tuck-in: Alloy Enterprises (2025) — proprietary thermal/heat-transfer technology for chillers, CDUs, and cold plates.

Operational:

• Business-system rollout: ~1,400 colleagues engaged, ~1,000 leaders trained, 150+ kaizens across ~20 priority areas; 7 “lighthouse” sites planned. Management is explicit that meaningful P&L impact is a 1–2-year story, not a single quarter.
• Margin inflection: adjusted EBIT margin to 15.5% (Q2 FY2026); FY2026 operating-leverage guide ~50%.
• Capacity ramp: North-American factory capacity built pre-Weidemanis is now ramping (a near-term productivity headwind that becomes a tailwind as it matures); management says capacity is sufficient for ~12–18 months of the order surge.

Headwinds:

• Middle East conflict: ~2–3% of company revenue, ~10%+ of EMEA; ~1/3 of the Middle East business was disrupted/delayed in Q2 FY2026.
• China / APAC: order growth has bottomed (~4% off an easy compare) but remains sluggish ex-data-center/semi/biologics.
• Security service repricing: management is deliberately trading volume for price in the less-differentiated security-service book — a near-term revenue drag (margin-accretive).
• 2025 cyclical softness in some short-cycle/residential-adjacent areas and governmental/institutional spending uncertainty.
• Reported asset-sale speculation: management was asked on the Q2 call about further portfolio moves and declined to confirm, framing the portfolio as “a sports team” still under review — leaving open further restructuring (and dilution/dis-synergy risk).

Verdict: The changes strengthen the thesis — a cleaner portfolio, a credible new operating leader, and a visible margin inflection. The headwinds are real but mostly manageable/transitory. The net effect over two years is a clearly better-positioned company; the open question is durability, not direction.

9. Risk Analysis (Risk Matrix)

Catastrophic-loss risk is low — JCI is a scaled, cash-generative, investment-grade franchise with a moderate balance sheet; the realistic downside is a 25–40% de-rating, not impairment of the enterprise. The fat-tail is the AFFF/PFAS litigation, which is genuinely open-ended but more likely to be a multi-hundred-million-to-low-billions settlement drag than an existential event.

10. Valuation Discussion (Embedded Expectations)

No price target and no recommendation. This section frames what the current price implies and the scenarios around it.

Where the multiple sits. At $144.96, JCI trades at:

• ~30x forward adjusted EPS ($4.85 FY2026 guide) — versus its own ~17–22x historical range and roughly level with Trane (~31x), a higher-margin/higher-ROIC business.
• ~22x forward EV/EBITDA (~27x trailing on FY2025 EBITDA), versus a low-teens-to-mid-teens historical range.
• ~3.4x EV/sales — near the top of its own history.
• ~2.4% free-cash-flow yield (continuing FCF ~$2.1B on ~$88.6B market cap).
• AZI own-history percentiles: composite 77.5th; price-to-sales 98.7th; price-to-book 98.3rd (the P/E percentile of 35.6th is an artifact of trailing GAAP EPS being inflated by the Bosch gain — ignore it; the P/S and P/B percentiles are the honest tell). On the metrics that aren’t distorted, JCI is the most expensive it has been in a decade.

What the price embeds. To justify ~30x forward earnings, the market must be underwriting some combination of: (1) the Danaher-style transformation substantially closing the margin/ROIC gap to Trane (segment margins from 17% toward 20%+, ROIC from ~10% toward high-teens), (2) the data-center order surge sustaining into a multi-year, structurally-higher growth rate rather than normalizing, and (3) continued double-digit adjusted-EPS compounding (margin expansion + buyback). That is a base-to-bull set of assumptions stacked on top of each other — each plausible individually, but the price affords little room for any to disappoint.

Scenario analysis (illustrative, not a target):

• Bull (~$175–$190): Business system delivers; FY2027–28 adjusted EPS compounds ~15–20% to ~$6.00+; data-center demand holds; margins march toward high-teens. A sustained ~28–31x on a much larger base. (Morgan Stanley’s $175 Overweight PT, raised from $140 on 2026-06-09, sits here.)
• Base (~$120–$140): EPS reaches the ~$4.85 guide and grows high-single-to-low-double-digit thereafter; the multiple normalizes modestly toward ~25–27x as the turnaround matures and the law of large numbers bites. Roughly flat-to-modestly-positive total return from spot — a “great direction, fully-priced” outcome.
• Bear (~$95–$110): Data-center orders normalize into a 2027 air-pocket; margins plateau near current levels; the multiple de-rates toward ~20x (still a premium) on a ~$5.00 number. A 25–35% drawdown — the risk of paying a quality multiple for quality-in-progress.

The reverse-DCF intuition: at ~30x forward and a ~2.4% FCF yield, the stock requires sustained low-double-digit FCF growth for a decade to deliver a market-beating return — achievable if the turnaround is structural and data-center demand durable, but demanding for a business that grew revenue ~6% organic and earns ~10% ROIC today.

Verdict: The valuation already capitalizes a successful, largely-completed turnaround and a durable data-center cycle. The market is underwriting the convergence-to-Trane and the cycle-as-secular correctly in direction but generously in degree — the embedded expectations are high enough that the risk/reward is asymmetric to the downside at this entry.

11. Variant Perception

Consensus view: JCI is a high-quality re-rating story — a “new Johnson Controls,” cleaned up into a pure-play commercial-buildings compounder, with a proven-operator CEO, a Danaher operating system inflecting margins, and a front-row seat to the data-center cooling super-cycle. Sell-side is constructive and raising targets (MS to $175). The bull narrative is that ~30x is justified by the EPS-growth trajectory (~30% in FY2026) and the long self-help runway.

The strongest bull case: This is a genuine operational transformation in its early innings. JCI has best-in-pool assets (Americas commercial/applied + a 32%-of-revenue service book + an integrated chiller franchise) earning sub-pool economics — the single largest self-help margin gap in the group. A credible Danaher operator closing even half that gap, against a multi-year data-center tailwind, supports years of mid-teens-plus EPS compounding. At that growth, ~30x is defensible and the stock keeps working.

The strongest bear case: The market is paying Trane’s multiple for Carrier’s-or-worse current returns, on the promise of convergence. The re-rate from ~$50 to ~$145 has already captured most of the turnaround in advance of the proof. The growth leg the bulls lean on (data centers) is exactly where the capital cycle is turning hostile and where order growth (+30–40%) is running 5x ahead of revenue (+6%), setting up 2027 normalization risk. Returns are only modestly above cost of capital, the comp plan ignores ROIC, the $5B buyback was executed at a record multiple, and the legacy AFFF/tax tails are non-trivial. Strip the one-time Bosch gain and the honest valuation is the most expensive JCI has ever been on sales — with no margin of safety.

The 3–5 assumptions that matter most:

1. Does the business system structurally lift margins toward high-teens/20% and ROIC out of the low teens? (The entire turnaround thesis.)
2. Is the data-center order surge durable, or a pull-forward that normalizes in 2027? (The growth thesis and the capital-cycle risk.)
3. Does the 32%-of-revenue service book monetize like Trane’s annuity (margin, attach, growth), or remain under-monetized?
4. Does the multiple hold near 30x, or mean-revert as the law of large numbers and cyclical normalization bite?
5. Do the AFFF/PFAS and Tyco-tax tails stay contained?

What would falsify each side:

• Falsifies the bull: two-plus quarters of segment-margin plateau near current levels while data-center book-to-bill rolls below 100% — convergence stalling against a top-of-group multiple.
• Falsifies the bear: sustained sequential margin/ROIC expansion toward high-teens with the backlog converting on schedule and service-attach growth accelerating — proof the Danaher system is structural, justifying the multiple on a larger base.

12. Fact vs. Interpretation Table

13. Open Questions

1. What is the actual service gross margin and attach rate? JCI discloses services at 32% of revenue but not the Trane-style economics (gross margin, attach, growth) needed to judge whether the annuity is best-in-class or merely large.
2. How much of the ~$20B backlog and the order surge is durable demand vs. customers ordering early (management acknowledged customers are placing orders earlier due to power/electrical-infrastructure lead times)?
3. What is the through-cycle ROIC target, and why is there no ROIC metric in the incentive plan for a company running the Danaher playbook?
4. Where does the portfolio review end? Management’s “sports team” framing and the asset-sale speculation leave open further divestitures — what, and at what dilution?
5. What is the realistic AFFF/PFAS personal-injury and state-AG exposure beyond the $750M water-systems settlement?
6. How quickly can EMEA close its ~300–400bps margin gap to regional peers, and is the catch-up structural or cyclical?
7. What is the normalized FY2027–28 EPS power once the buyback tailwind fades and data-center growth normalizes?

14. What Must Be True

For the bull case to be right:

1. The business system must structurally lift segment margins toward high-teens/20% and ROIC out of the low teens — and stay there through a cycle.
2. Data-center demand must prove a multi-year structural growth leg, not a 2025–26 pull-forward, with margins resisting the capital-cycle mean-reversion.
3. Adjusted EPS must keep compounding double-digit after the buyback tailwind fades, on organic margin + volume.

Falsification test: two-plus consecutive quarters of segment-margin plateau (no YoY expansion) and enterprise book-to-bill falling below 100% with data-center orders decelerating — the convergence stalling and the cycle rolling over together. If that happens against a ~30x multiple, the de-rating is the dominant outcome.

For the bear case to be right:

1. The multiple must mean-revert (from ~30x toward ~22–25x) as the turnaround matures, the law of large numbers bites, and the data-center cycle normalizes.
2. Margins must plateau near current levels while orders normalize into a 2027 revenue air-pocket.
3. The market must re-anchor on JCI’s actual ~10% ROIC and second-tier economics rather than the promise of convergence.

Falsification test: sustained sequential margin and ROIC expansion toward high-teens/low-20s with the ~$20B backlog converting on schedule and service growth accelerating — i.e., the Danaher system proving structural. If the returns visibly converge toward Trane’s, the premium multiple is justified on a much larger earnings base and the bear is wrong.

15. Source Appendix

(See Appendix B below for the full source catalogue. Primary sources first; all figures reconcile to these.)

Key primary sources: JCI FY2025 10-K (filed 2025-11-14, jci-20250930.htm) and FY2021–FY2024 10-Ks; the 2026 proxy statement (DEF 14A, 2026-01-16); the Q2 FY2026 earnings call transcript (2026-05-06). Quantitative figures cross-checked against public financial-data aggregators and SEC EDGAR XBRL; valuation own-history percentiles and the Morgan Stanley price-target raise (2026-06-09) from public market data. Peer context drawn from public filings and disclosures of Trane (TT), Carrier (CARR), Lennox (LII), Honeywell (HON), and Parker-Hannifin (PH).

This analysis takes no investment position and sets no price target. The single, clearly-labeled exception is the “Claude’s Take” block at the top, which is the author’s own independent subjective view. Every material claim above traces to the primary sources catalogued in the Source Appendix.

APPENDIX A — Standard Diligence Questionnaire

Johnson Controls International plc (NYSE: JCI) — as of 2026-06-13

Supplemental diligence questionnaire. Answers grounded in primary sources; Fact / Interpretation / Assumption labeled where it matters.

General

What thoughtful questions have other investors asked about this company? The recurring institutional questions (echoed on the Q2 FY2026 call): (1) Are data-center orders at a peak — “are we at peak orders?” (management: plateauing at 30–40% growth, pipeline still double-digit). (2) How long does the Danaher-style business system take to show up in the P&L (management: 1–2 years, “early innings”)? (3) How does content shift between chillers, air-handlers, and CDUs as liquid cooling scales (management: AHU content rising, chiller fears “overstated,” CDUs just ramping)? (4) What is the strategic endgame — further asset sales (management: portfolio “still under review”)? (5) Can the lower-margin/lower-ROIC business actually converge to Trane-like economics? The variant-perception debate is not on the business, but on the multiple and the durability of the data-center leg.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: Mid-cycle and rising, but with a cyclical asterisk. Margins are inflecting off a multi-year low (operating margin 7.8% in FY2020 → 12.0% FY2025 → 15.5% adjusted in Q2 FY2026), driven by internal self-help (the business system) more than the external cycle — which is bullish for durability. But the order surge (+30–40%) is partly cyclical/data-center-driven and runs well ahead of revenue (+6%), so the growth rate is closer to a high than a low.

Driven by external environment or internal actions? Both, but the margin expansion is predominantly internal (productivity, price/cost, 80/20, mix) — the higher-quality kind. The order/revenue acceleration is external (data-center capex super-cycle).

How stable are revenues? Moderately stable: ~32% services (recurring, multi-year contracts) provides a floor; ~68% install/project is longer-cycle and backlog-supported (record ~$20B backlog, ~70% converting within 12 months) but ultimately tied to commercial-construction and hyperscaler capex cycles.

Outlook for products/services? Strong near-term: ~6% organic growth, ~50% operating leverage, ~$4.85 adjusted EPS (~+30%) guided for FY2026. Medium-term hinges on data-center durability and the service-attach build-out.

How big is this market — growing/shrinking, domestic/international? Large and growing. Commercial/applied HVAC + fire + security + controls is a multi-hundred-billion-dollar global market; the data-center cooling sub-segment alone is sized ~$26B (2025) → ~$128B (2033, ~22% CAGR). JCI is ~56% US / ~44% international, ~150 countries.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? Stable-to-intensifying at the growth edge. The high-end commercial/applied oligopoly (Trane, Carrier, JCI, Daikin, Mitsubishi, Bosch) is rational, but data-center cooling is drawing a flood of capital (Vertiv, Schneider, Carrier, AAON, liquid-cooling specialists) — a Marathon capital-cycle warning that scarcity margins there will compress.

How profitable is the business (ROIC, ROE)? Fact: ROIC ~10% (FY2025), up from 5.7% (FY2020) — improving but only modestly above cost of capital, and ~half of Trane’s ~28%. ROE is unusable (~764% reported, an artifact of ASR-shrunken equity). Gross margin 36.4%, segment EBITA margin 17.0%.

How profitable is the industry — competitors, barriers? The profitable core (NA commercial/applied + service) earns 18–21% segment margins and high-teens-to-mid-20s ROIC for the best operators. Barriers are high: engineering depth, brand, code/efficiency regulation, installed-base service density, dealer/contractor lock-in.

Can the business be easily understood? Yes — it sells, installs, and services building HVAC/fire/security/controls. Simpler than the pre-2025 conglomerate.

Can it be undermined by foreign low-cost labor? Largely no for applied/engineered systems and field service (local, code-bound, relationship-driven). More exposed in commoditized product lines, but those are mostly divested.

Do brands matter? Yes — York (chillers), Metasys (controls), Simplex/Autocall (fire), Silent-Aire (data-center AHUs) carry real specifier/engineer credibility.

Nature of competition? Engineered-system performance, energy efficiency, application-specific reliability, installed-base service relationships, and increasingly the digital/controls layer (OpenBlue). Price matters more in commoditized security/products.

Customers’ switching costs? High for installed mission-critical systems (chiller plants, building-management, fire networks) — the basis of the service annuity and the moat.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The installed base and service relationships (an intangible annuity) are not capitalized; the OpenBlue software platform is largely expensed. Interpretation: genuine economic value not on the books.

Off-balance-sheet liabilities? Fact: The material ones — legacy asbestos (gross liability $387M, net ~$48M after insurance), environmental reserves ($187M, incl. PFAS sites), and open-ended AFFF/PFAS litigation (beyond the $750M water-systems settlement: 12,400+ individual/mass actions, 49 class actions, state-AG suits — unquantified). Plus $1.9B unrecognized tax benefits.

How conservative is the accounting? Reasonable, but watch the non-GAAP bridge: FY2025 reported EPS ($5.03) is inflated by a one-time Bosch gain; the company adjusts out ~$1.13/share of restructuring/amortization/transformation costs to reach $3.76 adjusted. Adjustments are explained but elevated. Goodwill ($16.6B) exceeds equity — impairment sensitivity.

How CapEx-hungry? Light — capex ~1.8% of sales (~$434M FY2025). Asset-light field-and-service model; ~100% FCF conversion guided.

Capital Allocation & Management

How much FCF, and how is it used? ~$2.1B continuing-ops FCF (FY2025). Policy historically ~100% returned via dividends + buybacks; now balanced with bolt-on tech M&A and balance-sheet discipline (~2x net leverage).

Significant acquisitions recently? Net divestiture (Bosch R&LC HVAC, ADTi, ADT Mexico) — value-rational simplification. Small tuck-in: Alloy Enterprises (thermal/heat-transfer tech).

Buying back shares? Yes, aggressively — $5.99B in FY2025 (the $5B ASR funded by Bosch proceeds), reducing the count ~7% (693M → 612M). Interpretation: defensible use of proceeds but executed near a record multiple (98th-pctile P/S) — flatters per-share metrics, questionable value.

Issuing large amounts of stock to insiders? No — SBC is low (~$140M, <1% of revenue); count is falling.

Compensation policy / motivations of management? Fact: Annual bonus on EBIT growth / revenue growth / FCF conversion (equally weighted); PSUs on cumulative pre-tax earnings / recurring revenue / relative TSR. No ROIC metric anywhere — a governance weakness for a capital-allocation-intensive, Danaher-style turnaround. CEO Weidemanis (ex-Danaher) got a $5M equity inducement (alignment-positive); insiders own <1% (alignment via grants, not stakes). FY2025 bonus paid at 176.6% of target.

Valuation & Market Data

ADR, MLP, or K-1 issuer? None — JCI is an Irish-domiciled plc common share trading on the NYSE (ordinary shares, not an ADR). No K-1. Tax note: Irish domicile creates the inversion/treaty risk and the ~$1.9B unrecognized-tax-benefit exposure; dividends to US holders are generally qualified, with Irish withholding considerations.

Dividend policy? Modest: ~$1.51/share (FY2025), ~+2%/yr, ~30–40% payout on adjusted continuing EPS, ~1.0% yield. Secondary to buybacks.

How profitable? ~10% ROIC, 36.4% gross / 17.0% segment-EBITA margin — improving, second-tier within the peer group.

Is net income diverging from cash from operations? FY2025 reported OCF (~$1.4B all-in) is below net income due to ~$1.16B of discontinued-ops outflows and working-capital build; continuing-ops OCF (~$2.55B) tracks adjusted earnings well. Normalize for discontinued ops before drawing conclusions.

Risks & Downside

What would cause the stock to decline? (1) Multiple de-rating from ~30x (the dominant risk — 98th-pctile P/S, near Trane’s multiple with half the ROIC); (2) data-center order normalization / 2027 air-pocket; (3) margin/ROIC convergence stalling; (4) an adverse AFFF/PFAS or tax development; (5) a cyclical commercial-construction downturn.

Risk of catastrophic loss? Low — scaled, investment-grade, cash-generative, moderate leverage. Realistic downside is a 25–40% de-rating, not enterprise impairment.

Chance of total loss? Negligible. The fat-tail (AFFF/PFAS) is a settlement-drag risk, not existential.

Recent News & Events

Has the business environment changed recently? Yes, materially over 24 months — new CEO (Mar 2025), Bosch divestiture (Jul 2025), resegmentation (Apr 2025), data-center demand surge, and a visible margin inflection. The near-term tape is quiet/positive (Morgan Stanley raised its PT to $175 on 2026-06-09; Q2 FY2026 beat-and-raise on 2026-05-06).

Significant acquisitions? Alloy Enterprises (thermal tech tuck-in, 2025); otherwise net divestiture.

Change in accounting policies? Resegmentation to three regional segments (Apr 2025); R&LC HVAC reclassified to discontinued operations. No adverse policy changes flagged.

Recent changes — new markets, facilities, management? New CEO and operating model; North-American capacity ramp; data-center “AI factory” reference designs; CDU/liquid-cooling entry; Aarhus (Denmark) heat-pump/chiller capacity expansion.

APPENDIX B — Source Appendix

Johnson Controls International plc (NYSE: JCI)

Primary sources first. All figures reconcile to these. Accessed 2026-06-13 unless noted. All sources below are public.

1. SEC Filings — Primary (EDGAR, CIK 0000833444)

Note on filing dates: EDGAR file dates reflect filing, not period-end. JCI’s fiscal year ends September 30; FY2025 = year ended 2025-09-30.

2. SEC XBRL Financial Facts (EDGAR, reconciled to 10-K)

• Continuing revenue: FY2020 $22,317M → FY2025 $23,596M (with portfolio noise from divestitures/resegmentation).
• Operating income (GAAP): FY2020 $1,746M → FY2025 $2,828M; operating margin 7.8% → 12.0%.
• Gross margin: 33.2% (FY2020) → 36.4% (FY2025).
• Diluted EPS: continuing $2.63, discontinued $2.40, total $5.03 (FY2025); adjusted continuing $3.76.
• Net debt $9.5B; total debt $9.9B; cash $0.4B; goodwill $16.6B; equity $12.9B.
• Dividends paid / buybacks / capex by year (FY2021–FY2025) per §7 table.
• ROIC ~10.05% (FY2025); ROE not meaningful (ASR-distorted).

3. Earnings Call / Management Commentary (hypothesis, validated vs filings)

• JCI Q2 FY2026 earnings call transcript — 2026-05-06 (public transcript / company IR). CEO Joakim Weidemanis, CFO Marc Vandiepenbeeck. Source for: orders +30%, backlog ~$20B (+26%), adjusted EBIT margin 15.5% (+310bps), adjusted EPS $1.19 (+45%), FY2026 guide raised to ~$4.85, ~6% organic / ~50% operating leverage, data-center commentary (York chillers, Silent-Aire AHUs, CDUs ~$100M, Alloy acquisition), business-system rollout, security-service repricing, Middle East impact.
• Historical JCI earnings call transcripts FY2022–FY2024 (George Oliver / Olivier Leonetti era; public company IR / transcript archives) — context for prior strategy (service transformation, ~100%-FCF-return policy, $230M productivity, Sept-2023 cyber incident, $12.1B FY2023 backlog).

4. Market / Quantitative Data Helpers (refreshed 2026-06-12/13)

• Financial-data aggregators — multi-period income statement, balance sheet, and cash flow (FY2020–FY2025), profitability ratios, enterprise value, and valuation multiples. Source for ROIC, EV (~$81B at FY2025 base; ~$98B at current price), margins, and multiples. Third-party aggregated; reconciled to filings.
• Own-history valuation percentiles (as of 2026-06-12): composite 77.5th percentile of JCI’s own ~10-year range; P/E 35.6th (distorted by the one-time Bosch gain in trailing GAAP EPS), P/B 98.3rd, P/S 98.7th; latest price $144.96. Sell-side: Morgan Stanley maintained Overweight and raised its price target $140→$175 (2026-06-09).
• SEC EDGAR (CIK 0000833444) and public market-data sources — price and share-count cross-checks.

5. Industry & Peer Cross-Reads

• Public filings and disclosures of peers — Trane Technologies (TT), Carrier (CARR), Lennox (LII), Honeywell (HON), and Parker-Hannifin (PH) — used for industry structure, peer valuation, data-center cooling sizing, and competitive positioning. All drawn from public 10-Ks, earnings materials, and market data.
• Third-party data-center cooling TAM and hyperscaler-capex figures triangulated from the peer reports (TT, CARR) and management commentary; treated as directional industry framing.

6. Key reconciliation notes

• GAAP-total vs continuing vs adjusted EPS: FY2025 $5.03 total is inflated by the ~$1.5B after-tax Bosch gain (in discontinued ops); the run-rate base is $2.63 GAAP continuing / $3.76 adjusted continuing → ~$4.85 FY2026 guide.
• Backlog: FY2025 10-K “Building Solutions” backlog $14.9B and RPO $22.7B vs management’s ~$20B Q2-FY2026 “record backlog” — a definitional step-up; both cited, directionally a genuine record.
• EV: ROIC reports ~$81B on the FY2025 (Sep-2025) market-cap base; at the 2026-06-12 price (~$145, ~611M shares) market cap is ~$88.6B and EV ~$98B — the memo uses the current figures for valuation.
• ROE / book-value ratios are distorted by the ASR-shrunken equity and goodwill > equity; ROIC and EV-based multiples are the reliable lenses.